Beruflich Dokumente
Kultur Dokumente
··································································································································································································································
Abstract
The design of stress-ribbon footbridges needs more iterations than the design of general structures. This is because, in addition to
the cross section of deck, the cross sectional areas of bearing and post-tensioning cables and the post-tensioning force should be
assumed. To minimize the iterations in this study, the regression equations of design variables, which can reasonably assume the
cross sectional area of bearing and post-tensioning cables and the post-tensioning force, are determined for various bridge lengths
with varying sag/span ratios. Also, the maximum values for the cable allowable stresses, assumed to determine the cross sectional
area of cables, are presented for various bridge lengths with various sag/span ratios. It is demonstrated through the design examples
that the suggested preliminary design method assuming design variables is efficient in minimizing iterations of the design process for
single-span and multi-span stress-ribbon footbridges and in determining the cross sectional areas of cables and the post-tensioning
force economically.
Keywords: stress-ribbon, bearing cable, post-tensioning cable, sag/span ratio, maximum cable stress, concrete stress at the top fibers
··································································································································································································································
*Member, Ph.D. Student, Dept. of Civil & Environmental Engineering, Kongju National University, Cheonan 331-717, Korea (E-mail: han_kj@kongju.ac.kr)
**Member, Professor, Dept. of Civil Engineering, Chungnam National University, Daejeon 305-764, Korea (E-mail: nhrim@cnu.ac.kr)
***Member, Professor, Dept. of Civil & Environmental Engineering, Kongju National University, Cheonan 331-717, Korea (E-mail: mgko@kongju.ac.kr)
****Member, Professor, Dept. of Civil & Environmental Engineering, Kongju National University, Cheonan 331-717, Korea (Corresponding Author, E-
mail: kkkim@kongju.ac.kr)
−1−
Ki-Jang Han, Nam-Hyoung Lim, Man-Gi Ko, and Kee-Dong Kim
As the sag ratio increases, the vibration characteristics of 2.2 Governing Limit State
stress-ribbon footbridges tend to be better because the anti- Stress-ribbon footbridges produce the required stiffness by
symmetric mode becomes the lowest mode and the natural introducing the post-tensioning force to the concrete deck with a
frequency grows away gradually from the frequency range of specific sag, but they show more flexible behaviors than common
1.5 hz to 2.3 hz for which the design criteria for pedestrian girder bridge because of the very thin deck compared to the
bridges (JPCEA, 2000) limit (Kunitomo et al., 1990). length of the bridges. Accordingly, the examination for the limit
Therefore, as the sag ratio is increased, it is quite possible to states of stress-ribbon footbridges is governed by the limit state
make the design economical and efficient, but the serviceability of serviceability. Stress-ribbon footbridges can be classified with
of pedestrians can also be dropped rapidly in case of the high sag two types of structures; one as Pre-stressed Concrete (PC) not
ratio. In the design criteria of footbridges, the maximum allowing the tensile stress on the concrete deck by service loads
longitudinal slope of footbridges is defined according to the and the other as Partially Pre-stressed Concrete (PPC) allowing
construction site and the purpose of use; less than 12% ( f/L = 1/ the tensile stress. Examination on the limit state of servi-
33) for the general use and less than 8% (f/L = 1/48) for disabled ceability is made by comparing the maximum tensile stress of
persons. Considering the effect of sag ratio on the behavior and the concrete deck to the tensile stress of the targeted value. In
the serviceability of footbridges, the sag ratio ranging from 1/50 Table 1, the various service load cases for stress-ribbon bridges
to 1/30 is generally used. When the weighting value is put on the are shown, and the stresses at the top fibers and the sags for
serviceability, the low sag ratio should be selected if possible, service load case 12 producing the maximum concrete stress
while the higher one should be selected reversely if the are presented in Fig. 1. Deck self-weight D1 shown in Fig. 1 acts
economical efficiency matters. on the bearing cables, and pre-stressing force PS is introduced
The cross-sectional area of deck needs to be as small as into the transformed section of the bearing cable cross-section
possible because the increase of cross-sectional area leads to the and the deck cross-section. Service loads such as pavement load
self weight increase and in turn the increase of horizontal reaction D2, creep and shrinkage (CR + SH), etc., being applied after the
and a larger substructure to sustain the increased horizontal introduction of post-tensioning force, are applied to the
reaction. In general, the limit states of serviceability and strength transformed section of the deck cross-section, the bearing cable
can be easily satisfied when the post-tensioning force is introduced cross-section, and the post-tensioning cable cross-section. The
such that enough stiffness guaranteeing both comfortable variation of sag according to the construction sequence corres-
walking and the stability of the shape can be developed even to ponding to each service load is shown in Fig. 1. Service loads
the deck with the very small depth determined by the minimum such as pre-stressing force PS, creep and shrinkage (CR + SH),
cover requirement of both cables and reinforcing bars (Strasky, and temperature decrease T (−) resulted in an upward deflection.
2006; Schlaich, J, 1996). From 108 stress-ribbon bridges built in In case of the design of PC structure in which the tensile stress
Japan from 1965 to 2011, it is evident that the depth of a deck is is not allowed, the maximum stress σtM of service load case 12
around 20 cm regardless of the bridge length. The width of a needs to be equal to or larger than σtM2 as shown in Fig. 1. To be
deck should be determined in consideration of the expected such, the initial stress at the top fibers σtI affecting σtM should be
density of pedestrians and local conditions, but the design will be equal to or larger than σtI2. σtI is the initial stress at the top fibers
economical when the width is as small as possible. produced by the introduction of the post-tensioning force and
Cable analyses of stress-ribbon bridges are more convenient and can be determined by Eq. (1) suggested by the reference
faster than finite element analyses. For stress-ribbon bridges that (JPCEA, 2000). In case of the design of PPC structure, σtI occurs
have a high flexural rigidity a large degree of end restraint, however,
the bending moment of deck obtained by the cable analysis can be
underestimated. In case that the ratio of moment of inertia over the Table 1. Service Load Cases
bridge length (I/L) is larger than 2.0 × 10−5, it is reported that there is Load Case No.
a big difference between the results by the cable analysis and by the 1 Deck construction (D1)
finite element analysis(Sakanishi et al., 2003). Therefore, if the 2 D1 + Prestressing (PS)
bridge length is short or the flexural rigidity is big, the structural 3 D1 + PS + Relaxation (R)
analysis of stress-ribbon bridges can be performed through the finite 4 D1 + PS + R + Pavement (D2)
element analysis using computer programs such as DIANA (TNO, 5 D1 + PS + R + D2 + Creep&Shrinkage (CR +SH)
2012) and RM Bridge (Bentley, 2012), etc. 6 D1 + PS + R + D2 + CR+SH + Live (L)
By contrast with other bridge structures, the sag of stress- 7 D1 + PS + R + D2 + CR + SH + Temperature (−)(T(−))
ribbon footbridges due to creep and shrinkage decreases 8 D1 + PS + R + D2 + CR + SH + Temperature (−)(T(+))
because the creep effect lifts themselves. In this study, the C-S 9 D1 + PS + R + D2 + CR+SH +Snow (SW)
model of JSCE (JSCE, 2013) and CEB-FIP Model Code90 10 D1 + PS + R + D2 + CR + SH + L + T(−)
(CEB, 1992) are applied to cable analysis models (Strasky, 11 D1 + PS + R + D2 + CR + SH + L + T(+)
2006; Bentley, 2012) in order to define the behavior of creep 12 D1 + PS + R + D2 + CR+SH + L + T(−) + SW
and shrinkage. 13 D1 + PS + R + D2 + CR + SH + L + T(+) + SW
80 m and a sag ratio of 1/40, were shown in Figs. 4 and 5, the post-tensioning force Pt can be calculated and the cross-
respectively. Since fD is defined by the sum of ffT and ∆sag as sectional area of post-tensioning cables Ap2 also can be determined
shown in Fig. 2, it was determined by changing ∆sag for ffT of 2 m. from Pt/fmax. fmax is cable allowable stress assumed by a designer.
The range of ∆sag used is 0.14 ffT ~ 0.26 ffT. The corresponding If the design is performed by using the cross-sectional areas of
value of fP for the change of fD was determined so that the final bearing and post-tensioning cables and the post-tensioning force
sag ff , obtained by service loads up to creep/shrinkage, matches determined as above, the limit states of serviceability and
up with the final targeted sag ffT. The stress at the top fibers of the strength as well as the final targeted sag can be satisfied without
deck according to the change of ∆sag/ ffT (∆sag= fD − ffT) was shown iterations. In order to apply this process to stress-ribbon footbridges
in Fig. 4. If the target σtM is placed on the solid line of the graph, with the generally used length ranging from 30 m to 150 m and
like dot A of Fig. 4, ∆sag/ ffT and σtI (dot A’) corresponding to the various sag ratios, ∆sag/ffT- σtM relationship (Line 1), ∆sag/ffT - σtI
target σtM can be determined directly from that solid line of the relationship (Line 2), and ∆sag/ffT - fP relationship (Line 3) as
graph. In case that the target σtM isn't placed on the solid line, like shown in Figs. 4 and 5 were determined for the total eighteen
dot B, ∆sag/ffT of dot B can be determined by the linear stress-ribbon footbridges with the length of 30 m, 40 m, 50 m,
interpolation of dot A and dot C, and ótI of dot B by the linear 60 m, 80 m, and 120 m applying the sag ratio of 1/30, 1/40, and
interpolation of dot A' and dot C', respectively. The sag of load 1/50 respectively, and the regression analysis was performed.
case 12 corresponding to ∆sag/ffT, determined from Fig. 4, can be
decided like the dotted line k obtained by the linear interpolation 2.4 Assumption for Cross-Sectional Areas of Cables and
of the solid lines in Fig. 5. The y coordinates of dot E’ and dot E Post-Tensioning Force
on the dotted line k indicate fD and fP, respectively.
The horizontal reaction H0 (= wd l2/8fD) due to the deck self 2.4.1 Assumption for Governing Design Variables Using
weight can be estimated by using fD of Fig. 5, and the cross- Regression Equations
sectional area of bearing cables Ap1 also can be determined from The results of regression analysis for ∆sag/ffT- σtM relationship,
H0/fmax. If fD, fP, and σtI of Figs. 4 and 5 are substituted for Eq. (1), ∆sag/ffT - σcI relationship, and ∆sag/ffT - fP relationship determined
from the design of the total eighteen footbridges were shown in
Figs. 6, 7, and 8, and Table 2. In Figs. 6, 7, and 8, the result of
regression analysis for the bridge length of 80 m was presented.
For the bridge lengths of 30 m, 40 m, 50 m, 60 m and 120 m, the
result of regression analysis was presented in Table 2. From the
results of regression analysis in Figs. 6, 7 and 8, it can be seen
that the three relationships above can be defined appropriately as
a linear form. The determination process of governing design
variables fD, σcI, and fP through the regression equations of Figs. 6, 7,
and 8, is as follow. First step is to determine ∆sag/ffT by the
substitution of target σtM for regression equations of Fig. 6, and
the next is to decide σcI and fP through the substitution of
determined ∆sag/ffT or regression equations of Figs. 7 and 8. For
example, in case of limiting the maximum concrete tensile stress
σtM to 0.0 MPa for PC structure, x value (x coordinate of dot A)
Fig. 4. Stress at The Top Fibers According to ∆sag/ffT Variation
obtained by substituting the target σtM of 0.0 MPa (y value of dot design variables determined from the regression equations of
A) for the regression equation of Fig. 6 is ∆sag/ffT which satisfies Table 1 will have an error because of the difference in both the
the final targeted sag ffT. Also, y values obtained by the substitution self-weight and the moment of inertia. To investigate the error of
of the decided ∆sag/ffT (x value of dot B and dot C) for the governing design variables when the cross-section is changed,
regression equations of Figs. 7 and 8, are σcI and fP, respec- ∆sag/ffT - σtM relationships are determined respectively by applying
tively. three different cross-sections to the stress-ribbon footbridge with
∆sag/ffT -span length and σcI -span length relationships determined the length of 80m and the sag ratio of 1/40. The first cross-
to satisfy the final sag ratios of 1/30, 1/40, and 1/50 and the section is the one shown in Fig. 3 (cross-section A). The second
targeted maximum stress at the top fibers of 0.0MPa are shown one has an effective width of 1.5 m and a depth of 0.2 m (cross-
in Fig. 9 and Fig. 10, respectively. As the span length decreases, section B; A = 0.5910 m2, I = 0.00337 m4), and the third one has
∆sag/ffT and σcI decrease linearly for the span length larger than an effective width of 2.0m and a depth of 0.18 m (cross-section
60m while ∆sag/ffT and ócI increase non-linearly for the span C; A = 0.6330 m2, I = 0.00301 m4). The results of regression
length less than 60m. The regression equations in Table 2 were analysis for ∆sag/ffT - σtM relationships of cross-sections A, B, and
determined at an interval of 10m for the span length less than C are presented in Fig. 11.
60 m to define the non-linear relationships of design variables Compared to cross-section A, cross-section B showed an
and span length. Linear interpolation of the regression equations increase of 8.8% and 23.9% in the cross-sectional area and the
shown in Table 2 can be used for span lengths that the regression moment of inertia respectively, while cross-section C showed an
equations are not presented. increase of 16.6% and 10.7% respectively. For cross-section B,
Regression equations of Table 2 are determined from using the the increase of the moment of inertia due to the increase of deck
cross-section configuration with the effective width of 1.5 m and depth occurs larger than the increase of bending moment due to
the deck depth of 0.18 m presented in Fig. 3. If the cross-section the increase of self-weight, while for cross-section C, the
different from that of Fig. 3 is used in the design, governing increase of bending moment due to the increase of self-weight
Table 3. Design Results of Exact and Approximated Solutions for Various Cross-sections
x-section B x-section C
B0 B1 C0 C1 C2
Max. Horizontal Reaction(kN) 9153 9124 10307 10258 10256
Post-tensioning Force(kN) 7461 7461 8577 7978 8469
fD (m) 2.458 2.464 2.482 2.464 2.486
ffT (m) 2.0 2.005 2.0 2.008 2.009
σtM (MPa) 0.0 0.056 0.0 -0.406 -0.009
Bearing Cable Area (mm2) 5825 5767 6127 6172 6117
Posttensioning Cable Area (mm2) 8983 8983 10365 9605 10227
cross-section C, it is shown that σtM of the approximate value is cables efficiently and satisfy the limit state of strength.
the unintended tensile stress of -0.406 MPa, and the cross- The design of the stress-ribbon footbridge of 80 m length with
sectional area of post-tensioning cables and post-tensioning force the cross-section of Fig. 3 and the sag ratio of 1/40 was
based on the approximate value are about 7% less than those performed in order to investigate the variance of (Sd/Rd) γi - σtM
based on the accurate one. For the case that σtM based on the relationship resulting from the change of fmax. The scope of fmax
approximate value differs significantly from the target σtM, the used is 0.45fpu ~ 0.70fpu. (Sd/Rd) γi was determined from changing
difference can be corrected through the iterations. The error the maximum stress at the top fibers σtM from -3 MPa to 1 MPa
correction of the approximate σtM of -0.406 MPa for cross-section for the respective fmax, and the results are shown as (Sd/Rd) γi -σtM
C, can be achieved through the calculation of σ of 0.406 MPa relationships for various fmax’s shown in Fig. 12.
(Fig. 11), obtained by subtracting the approximate σtM of -0.406 MPa Dot A and dot B of Fig. 12 are determined by aiming at σtM of -
from the target σtM of 0.0 MPa, and through the determination of 0.24 MPa and -3.0 MPa respectively and applying fmax of 0.55fpu.
the governing design variables by applying σ again to the Since the maximum stress at the top fibers of dot A is restricted
regression equations in Table 2. The maximum stress at the top more severely than that of dot B, the post-tensioning force of dot
fibers σtM re-determined through this process appears to be A is larger than that of dot B. (Sd/Rd) γi of dot A is smaller than
-0.009 MPa (dot C2 of Fig. 11), which is about the same as the that of dot B because the cross-sectional area of post-tensioning
target σtM, and the primary design results are very similar to those cables for dot A is larger than that for dot B.
based on the accurate value, as shown in Fig. 11 and Table 3. Dot C was determined by applying fmax of 0.7fpu and aiming at
σtM of -0.24 MPa. For the case of dot C, the post-tensioning force
2.4.2 Efficient Assumption for Cross-Sectional Areas of identical to dot A was introduced because σtM of dot C was the
Cables same as that of dot A. However, the cross-sectional areas of
The strength limit state of stress-ribbon footbridges is examined bearing and post-tensioning cables were decreased because of
by investigating the failure of cross-section as done in the case of the larger fmax than that of dot A and the value of (Sd/Rd) γi is
general members. Normally, examination of the cross-section bigger than that of dot A. For both dot A and dot C, the limit state
failure should be performed by checking the safety for bending of strength is satisfied and the horizontal reactions are almost the
moment, shearing force, torsional moment, and the axial force. same, but the larger fmax of dot C can be considered to be more
However, since the stress-ribbon footbridges resist to the efficient because the cross-sectional areas of cables for dot C are
external force mainly with the axial force, the examination on the smaller than those for dot A.
strength limit state can only be performed by checking the safety For the case of dot C, the cross-sectional area of bearing cables
for the axial force with the exception of the haunches at piers and decreases due to the larger fmax than that of dot B, and the cross-
abutments (JPCEA, 2000). The failure of deck cross-section in sectional area of post-tensioning cables increases due to the
stress-ribbon footbridges generally arises when cables rupture much larger post-tensioning force than that of dot B. And the
after concrete cracks occur and reinforcing bars yield. Accordingly, total cross-sectional areas of bearing and post-tensioning cables
the examination of strength limit state for the cross section for dot C and dot B appear to be identical because of the same
failure is performed by comparing the maximum tension of value of (Sd/Rd) γi . Though the total cross-sectional areas of
cables with the design strength of cables like Eq. (2). In Eq. (2), cables and the horizontal reactions for dot C and dot B are almost
Sd is the maximum tension of cables due to the factored loads, Rd the same, fmax of 0.7fpu for dot C can be considered to be more
is the design strength of cables defined by the strength and the efficient than fmax of 0.55fpu for dot B because σtM for dot C is
cross-sectional area of cables, and ri is the load modifier. restricted more severely. Therefore, σtM and fmax of dot C in Fig. 9
Therefore, the cross-sectional area of cables becomes the
minimum when being determined so that the value of (Sd /Rd) γi
can be 1.0.
( Sd /Rd )γi ≤ 1.0 (2)
In general, the maximum stresses of bearing and post-
tensioning cables occur from the installation of deck and the
introduction of post-tensioning force, respectively. Accordingly,
the cross-sectional areas of bearing and post-tensioning cables
are determined from AP1 = H0/fmax and AP2 = Pt /fmax , respectively.
fmax is the cable allowable stress assumed by a designer and is
generally regulated not to be greater than 0.7fpu, and fpu is the
ultimate strength of cables. Since the cross-sectional area of
cables depends on the value of fmax, σtM -(Sd /Rd) γi relationships
according to the change of fmax are investigated to determine the
appropriate fmax which can reduce the cross-sectional area of Fig. 12. σtM -(Sd/Rd)γi Relationships for Various fmax
are the most economical and the most efficient for the domain lengths less than or equal to 80 m except for the bridge with the
that the value of (Sd /Rd)ri and σtM are less than and equal to 1 and length of 80m and the sag ratio of 1/30, the maximum fmax is
σ ctM, respectively. For σtM between σ ctM and 0.0 MPa in Fig. 9, 0.7fpu and the cross-sectional area of cables can be decreased
fmax of 0.7fpu (Line C-D) can decrease the cross-sectional area of most efficiently by using the maximum fmax of 0.7fpu for the range
cables most efficiently. The selection of the appropriate σtM on of σtM defined by Line C-D of Fig. 13. If the tensile stress is
Line C-D can be made by considering cost-benefit because the restricted from PPC structure of dot C to PC structure of dot D,
cross-sectional areas of cables for dot C is smaller than that for the cross-sectional area of cables increases by 2 ~ 9% for the
dot D whereas σtM for dot D is restricted more severely.(because bridge of 80 m length, whereas it increases by about 43 ~ 68%
the more severely σtM is restricted, the more significantly the for the bridge of 40 m length. Therefore, it should be noted that
cross-sectional area of cables increases). the difference of cable areas used for PPC structure and PC
(Sd/Rd)γi - σtM relationships for the maximum fmax, according to structure can be quite large for short bridges. The allowable
the bridge length and the sag ratio, are shown in Fig. 13. The tensile stress of concrete σall, shown in Fig. 13, was determined
maximum fmax is the maximum of cable allowable stress, which by using the concrete strength of 50 MPa. For the bridge lengths
can reduce the cross-sectional area of cables to a minimum, and the sag ratios that the maximum fmax is not provided, linear
satisfying the limit state of strength. As the bridge length and the interpolation of the maximum fmax presented in Table 4 can be
sag ratio increases, the maximum fmax decreases. For the used.
footbridge with the bridge length of 80m and the sag ratio of 1/
30, the maximum fmax appeared to be less than 0.7fpu. For the 3. Design Examples
bridge lengths of 100 m and 120 m, the maximum fmax decreases
more and more as the sag ratio increases, as shown in Table 4. The design of total two footbridges of a single-span and a two-
For the case with the bridge length and the sag ratio where the span with the sag ratio of 1/30 shown in Fig. 14 was performed
maximum fmax appears to be less than 0.7fpu, σtM of 0.0MPa for using the regression equations of Table 2. To design them as a
dot E in Fig. 13 and the corresponding maximum fmax in Table 4 PC structure not allowing the tensile stress on the concrete deck,
will produce the most efficient design results. For the bridge the maximum stress at the top fibers σtM was required not to be
less than 0.0 MPa. And the deck cross-section shown in Fig. 3
was used.
Governing design variables ∆sag/ffT, σcI, and fp for the length of
100 m, determined from the linear interpolation of the regression
equations for the lengths of 80 m and 120 m in Table 2, are
shown in Table 5. The maximum fmax for the length of 100 m in
Table 4 is 0.64fpu but the cross-sectional areas of bearing and
post-tensioning cables are determined through the application of
fmax of 0.60fpu while considering the factor of safety, and the
results are shown in Table 5. By using the design variables and
the cross-sectional areas of cables in Tables 5, structural analysis
was performed and the concrete stresses at the top fibers and the
sags produced by various design loads are shown in Fig. 15(a).
The final sag ff became identical to the final targeted sag ffT
without iterations by using the regression equations presented in
Fig. 13. σtM - (Sd/Rd) γi Relationships for Maximum fmax of Various Table 2, and the targeted PC structure was accomplished as the
Footbridges maximum stress at the top fibers σtM appeared to be 0.034 MPa.
Table 4. Maximum for fmax Various Bridge Lengths and Sag Ratios
L = 80 m L = 100 m L = 120 m
f/L 1/50 1/40 1/30 1/50 1/40 1/30 1/50 1/40 1/30
Max, fmax 0.7 0.7 0.67 0.69 0.68 0.64 0.69 0.66 0.61
Table 5. Determination of Design Variables and Cable Cross-sectional Areas for a Single-span Footbridge
d = 180 mm d = 160 mm
fmax fmax = 0.60fpu fmax = 0.64fpu From Table 4
− Select σtM = 0.0 MPa − Select σtM = 0.26 MPa
− Determine ∆sag/ffT from Table 2 (σtM) − Determine ∆sag/ffT from Table 2 (σtM)
0.0 = 65.5x − 6.8 ⇒ ∆sag/ffT (80 m) = 0.104 0.26 = 65.5x − 6.8 ⇒ ∆sag/ffT (80 m) = 0.108
∆sag/ffT 0.0 = 70.7x − 8.3 ⇒ ∆sag/ffT (120 m) = 0.117 0.26 = 70.7x − 8.3 ⇒ ∆sag/ffT (120 m) = 0.121
Linear interpolation ⇒ ∆sag/ffT (100 m) = 0.111 Linear interpolation ⇒ ∆sag/ffT (100 m) = 0.115
− Calculate fD − Calculate fD
fD (100 m) = 0.111 × 3.333 m + 3.333 m = 3.703 m fD (100 m) = 0.115 × 3.333 m + 3.333 m = 3.716 m
− Get σcI from Table 2 (σcI) − Get σcI from Table 2 (σcI)
y = 102.3 × 0.104-2.7 ⇒ σcI (80 m) = 7.939 MPa y = 102.3 × 0.108-2.7 ⇒ σcI (80 m) = 8.348 MPa
σcI
y = 107.0 × 0.117 − 2.7 ⇒ σcI (120 m) = 9.819 MPa y = 107.0 × 0.121-2.7 ⇒ σcI (120 m) = 10.247 MPa
Linear interpolation ⇒ σcI (100 m) = 8.879 MPa Linear interpolation ⇒ σcI (100 m) = 9.298 MPa
− Get fp from Table 2 ( fp) Get fp from Table 2 (fp)
y = 1.3 × 0.104 + 2.70 ⇒ fp (80 m) = 2.835 y = 1.3 × 0.108+2.70 ⇒ fp(80 m) = 2.840
fp
y = 2.0 × 0.117 + 4.04 ⇒ fp (120 m) = 4.274 y = 2.0 × 0.121+4.04 ⇒ fp(120 m) = 4.282
Linear interpolation ⇒ fp (100 m) = 3.555 m Linear interpolation ⇒ fp(100 m) = 3.561 m
− Horizontal force (H0) − Horizontal force (H0)
wd × L2 13.575 × 1002 wd1 × L2 12.375 × 1002
H0 = ----------------- = -------------------------------- = 4282.4 kN H0 = ------------------
- = ------------------------------ = 4163 kN
8 × fD 8 × 3.703 8 × fD 8 × 3.716
Bearing Cable
− Required Cable Area (ΣAp1) − Required Cable Area (ΣAp1)
H 0 4582400 H 0 4163000
ΣAp1 = ---------------------- = --------------------------- = 4128 mm2 ΣAp1 = ---------------------- = --------------------------- = 3516 mm2
0.60 × fput 0.60 × 1850 0.64 × fput 0.64 × 1850
− Pre-stressing force (Pt) − Pre-stressing force (Pt)
wd1 × L ⎛ ----
2
1 1 wd × L
2
1 1
- × ⎝ - – -----⎞⎠ =5012 kN
pt = σcIAe1 + ------------------ - × ⎛ ----- – -----⎞ = 4784 kN
pt = σcIAe1 + ------------------
1
⎝ fP fD ⎠
8 fP fD 8
Post-tensioning Cable
− Required Cable Area (Ap2) − Required Cable Area (Ap2)
Pt 5012000 - Pt 5012000 -
ΣAp2 = ---------------------
- = -------------------------- = 4215 mm2 ΣAp2 = ---------------------
- = -------------------------- = 4041 mm2
0.60 × fput 0.60 × 1850 0.60 × fput 0.60 × 1850
The values of (Sd/Rd) γi and fmax/0.7fpu were 0.90 and 0.86, determined from using the governing design variables of Table 5
respectively, and the limit state of strength and the limit of fmax and the maximum fmax of 0.64fpu, the targeted value of 0.0 MPa
were satisfied. for σtM was not achieved. To correct the error, the governing
Considering that the cross-sectional areas of cables can be design variables were re-determined by applying σ of 0.26 MPa
decreased if the maximum fmax is used and that the effective to the regression equations of Table 2, and are shown in Table 5.
depth of deck shown in Fig. 3 can be decreased if the minimum Through the application of re-determined governing design
cover thickness is employed, the design was performed again by variables and the maximum fmax, the cross-sectional areas of
reducing the effective depth up to 0.16 m and by minimizing the bearing and post-tensioning cables were determined again and
cross-sectional areas of cables, are presented in Table 5. Structural analysis was performed by
As the maximum stress at the top fibers σtM was -0.26 MPa using the re-determined governing design variables and the
Fig. 15. Variation of Concrete Stresses and Sags According to Service Loads: (a) Single-span, (b) Two-span
cross-sectional areas of cables, and Fig. 15(a) shows the stresses structure was accomplished as the maximum stress at the top
at the top fibers and the sags due to various design loads. Even fibers σtM appeared to be 0.009 MPa. The values of (Sd /Rd) γi and
though there was a change in the cross-sectional area of deck, the fmax/0.7fpu were 0.83 and 0.99, respectively, and the limit state of
final sag ff became identical to the final targeted sag ffT with only strength and the limit of fmax were satisfied. From the comparison
one iteration, and the targeted PC structure was accomplished as of the two design examples, it can be shown that the bearing
the maximum stress at the top fibers σtM appeared to be cable area and the post-tensioning cable area of the two-span
0.045 MPa. Also, the limit state of strength and the limit of fmax bridge with the total length of 100m decrease by 57% and 23%,
were satisfied as the values of (Sd/Rd) γi and fmax/0.7fpu were 0.96 respectively, and the maximum horizontal reaction Hmax
and 0.89, respectively. decreases by 52%, as compared to the single-span footbridge of
By applying the deck depth of 0.16m and the maximum fmax of 100 m.
0.64fpu, the deck area decreases by 9%, the cross-sectional areas It is demonstrated through the design examples that the
of bearing and post-tensioning cables decrease by 7% and 14%, suggested preliminary design method assuming design variables
respectively, and the maximum horizontal reaction Hmax is efficient in minimizing iterations of the design process for
decreases by 6%, as compared to the first design. It has been single-span and multi-span stress-ribbon footbridges. The suggested
shown that the economical design can be performed by method can be effectively used to economically determine the
minimizing the deck cross-section if possible and by using the deck depth, the cross sectional areas of cables, and the post-
maximum fmax. tensioning force for single-span and multi-span stress-ribbon
The design for a total 100 m two-span bridge with the span footbridges with the generally used length ranging from 30 m to
length of 50 m was performed using the same cross sectional 150 m.
areas of cables and the same post-tensioning force for two spans.
Governing design variables ∆sag/ffT, σcI, and fp, determined from 4. Conclusion
the regression equations for the 50 m length of Table 2, are
shown in Table 6. The cross-sectional areas of bearing and post- Since the cross sectional areas of bearing and post-tensioning
tensioning cables were determined through the application of fmax cables and the post-tensioning force in addition to the cross
of 0.7fpu. By using the design variables and the cross-sectional section of deck should be assumed, the design of stress-ribbon
areas of cables in Tables 6, structural analysis was performed and footbridges needs more iteration processes than the design of
the concrete stresses at the top fibers and the sags produced by general structures. Therefore, in this study, it was intended to
various design loads are shown in Fig. 15(b). The final sag ff present the preliminary design method assuming design variables
became identical to the final targeted sag ffT without iterations by that can minimize iteration processes occurring in the design and
using the regression equations of Table 2, and the targeted PC can make the design efficient as possible. Regression equations
Table 6. Determination of Design Variables and Cable Cross-sectional Areas for a Two-span Footbridge
L = 50 m (1) L = 50 m (2)
fmax fmax = 0.70fpu fmax = 0.70fpu
− Select σtM = 0.0 MPa
− Determine ∆sag/ffT from Table 2 (σtM)
∆sag/ffT 0.0 = 56.3x − 5.8 ⇒ ∆sag/ffT (50 m) = 0.103 fD (50 m) = 0.103 × 1.67 m + 1.67 m = 1.842 m
− Calculate fD
fD (50 m) = 0.103 × 1.67 m + 1.67 m = 1.842 m
− Get scI from Table 2 (σcI)
σcI -
y = 102.6 × 0.103 − 3.4 ⇒ σcI (60 m) = 7.168 MPa
− Get fp from Table 2 ( fp)
fp -
y = 0.9 × 0.103 + 1.69 ⇒ fp(60 m) = 1.783
− Horizontal force (H0)
2
wd × L 13.575 × 50
2
for the various design variables that can assume efficiently the σcI = Initial stress at the centroid fibers
cross sectional area of bearing and post-tensioning cables and the σtI = Initial stress at top fibers
post-tensioning force were determined for the total eighteen σtM = Maximum stress at top fibers
footbridges with the bridge length of 30 m, 40 m, 50m, 60 m, ∆sag = Sag by fD plus the sag change by the introduction
80 m and 120 m applying the sag ratio of 1/50, 1/40, and 1/30 of post-tensioning force, relaxation, pavement and
respectively, and the process of utilizing those was presented. creep/shrinkage
Also, the maximum values for the cable stresses which are ∆Psag = Sag change by the introduction of post-tensioning
assumed for determining the cross-sectional areas of cables are force
determined for the total 12 footbridges with the bridge length of
40 m, 80 m, 100 m, and 120 m applying the sag ratio of 1/50, 1/ References
40. and 1/30 so that the cross-sectional areas of cables can be
minimized satisfying the strength limit state. Bentley (2012). RM Bridge V8i Analysis User Gulde, Bentley Systems,
It is demonstrated through the design examples that the Incorporated.
presented preliminary design method assuming design variables Cobo del Arco, D., Aparicio, A. C., and Mari, A. R. (2001). “Preliminary
design of pre-stressed concrete stress ribbon bridges.” ASCE
is efficient in minimizing iterations of the design process for
Journal of Bridge Engineering, Vol. 6, pp. 234-241, DOI: 10.1061/
single-span and multi-span stress-ribbon footbridges. The suggested (ASCE)1084-0702(2001)6:4(234).
preliminary design method assuming design variables can be Comite Euro-international du Beton (1992). CEB-FIB Model Code for
effectively used to economically determine the deck depth, the Concrete Structures, Cement and Concrete Association, Wexham
cross sectional areas of cables, and the post-tensioning force for Springs, Slough, England.
single-span and multi-span stress-ribbon footbridges with the Festschrift - Ulrich Finsterwalder. (1973) 50 Jahre für Dywidag, Verlag
length ranging from 30 m to 150 m. G. Braun, Karlsruhe.
Japan Pre-stressed Concrete Engineering Association. (2000). Standard
of Design and Construction of Stress-Ribbon Bridges, JPCEA, Japan
Notations (in Japanese).
Japan Society of Civil Engineers. (2013). Standard Specification for
Ae1 = Transformed area of deck and bearing cables Concrete Structures Design Operations, JSCE, Japan (in Japanese).
AP1 = Cross-sectional area of bearing cables Kalafatic, I.; Radic, J., and Medak, M. (2006). “Preliminary design
AP2 = Cross-sectional area of pre-tensioning cables procedure for one span post-tensioned stress-ribbon bridges.”
fD = Sag by the deck self weight DAAAM International Scientific Book, Vol. 25, pp. 313-328.
ff = Final sag Kunitomo, N. and Kumagai, S. (1990). “Design and construction of a
ffT = Targeted final sag stress-ribbon Bridges(1).” Journal of Pre-stressed Concrete, Japan,
Vol. 32, No. 2, pp. 71-75 (in Japanese).
fP = Sag by the introduction of the post-tensioning force
Meguru, T. and Yasuo, K. (1999). “Estimation of damping ratio of stress
and the deck self weight ribbon bridges an influence on serviceability.” JSCE, Nos. 612/I-46,
fmax = Cable allowable stress assumed by a designer pp. 337-348.
H0 = Horizontal reaction due to the deck self weight Sakanishi, K., Takuya, K., and Akio, S. (2003). “Comparison of finite
Hmax = Maximum horizontal reaction deformation theory and cable theory for stress-ribbon bridges.”
I= Moment of inertia Proceedings of the 59th Annual Conference of the Japan Society of
MPS = Moment by the introduction of the post-tensioning Civil Engineers, JSCE, Japan, IV-404, pp. 807-808 (in Japanese).
force Schlaich, J. (1996). “Stress ribbon concrete bridges.” Structural Engineering
International, Vol. 6, No. 4, pp. 271-274, DOI: http://dx.doi.org/
Pt = Post-tensioning force in which the instantaneous
10.2749/101686696780496247.
loss is reflected Strasky, J. (2006). Stress Ribbon and Cable-Supported Pedestrian Bridges,
wd1 = Deck self weight Thomas Telford Ltd, London.
yt = Distance from the centroid to the top fibers TNO (2012). TNO DIANA user’s manual, TNO Company.